Petroleum gas often contains a variety of acidic, gaseous contaminants, of which the principal ones are hydrogen sulfide, mercaptans and other diverse sulfur compounds, carbon dioxide, and carbonyl sulfide (COS). It is well known in the gas treating industry that such contaminants can be successfully removed by contacting the gas with aqueous solutions of one or more amines, which may be either selective or non-selective in their ability to absorb various of the acid gases. After such absorption, the acidic compounds are stripped from the amines and the amines are returned to the system, except to the extent they may have been lost in the process. It has been theorized that many different amines would provide some level of utility for removal of acid gases, but as a practical matter, the amines actually in commercial use are monoethanolamine (MEA), diethanolamine (DEA), methyldiethanolamine (MDEA), and diisopropanolamine (DIPA). Triethanolamine (TEA) is also frequently disclosed in the art as useful in gas treatment, but its actual commercial use appears to be very limited to non-existent. Use of MDEA/DIPA mixtures has also been reported (U.S. Pat. No. 4,808,765) for the purpose of removing H.sub.2 S and COS from liquefied petroleum gas (LPG). More specifically, U.S. Pat. No. 4,808,765 teaches that MDEA, a selective H.sub.2 S absorbant, may be formulated with DIPA, a COS absorbent, to reduce amine losses due to solubility in LPG. This patent also teaches that MDEA is less soluble than MEA or DEA in liquid hydrocarbons.
Treatment of LPG presents particular problems in that amines tend to be significantly soluble in the LPG, leading to a corresponding economic penalty due to the need to make up the lost amine(s). Many refineries use aqueous DIPA or MDEA to remove the acidic impurities from LPG; however, the concentration of these amines is typically limited to the range of about 20-35 weight percent of the aqueous stream in which they are supplied to the process. Operation at higher concentrations, which is desirable for capacity reasons, generally results in undesirably high levels of LPG contamination with amine(s). The problem is particularly acute at refineries treating cracked (i.e., highly unsaturated) LPG. Often, the loss rate of MDEA is sufficient to negate the economic justification for substituting MDEA for DEA. In addition to the high amine replacement costs, specialized remediation equipment is required, which increases the financial burden. Moreover, failure to remove dissolved MDEA can negatively affect downstream processes, e.g., poisoning of alkylation catalyst beds, and the like.
It would be highly desirable to have an amine composition which maximizes the effective amine concentration circulating in the LPG system, while yet minimizes the amount of amine(s) lost due to solubility in the LPG and increases desirable CO.sub.2 slip.